Generally Data warehouse is a database used for analysis and reporting. Data is uploaded on Database through different sources such as marketing & sales department. Data ware houses can be sub divided into Data mart. It stores the subset of data in warehouse. Different Data warehousing systems have different structures, some of the operational data store, while some have multiple data marts. Capacity of these warehousing may be different, some have dozens of data while some have a small number of data sources. In general, all data warehouse systems have the following structures-Data source layer, Data extraction layer, Staging area, ETL layer, Data storage layer, Data logic layer, Data presentation layer, Meta data layer, System operation layer. METADATA LAYER Data Data Staging E T L Data Data Data

Source Extraction Layer



Logic Presentation




Layer Layer

SYSTEM OPERATION LAYER Data source layer represent the different data sources, it can be any format such as word format, excel format or in relational database. Many different types of data can be data sources such as data from market research, data from operations such as sales data, marketing data, inventory management and product data.It also include third party data such as census data and demographic data Data Extraction Layer extract the data from data source to data warehousing, no major


transformation of data occur only cleansing of data may occur. Staging area transform the data into data marts, In E T L layer logic is applied on the data layer data become analytical in nature. In Data storage layer transformed data is stored, where in Data logic layer stores business rules. Data presentation layer contain the data that reaches the user, it can be in text form, tabular form or in presentation form. Meta data layer is where information about the data stored in data warehouse is stored. System operation layer store the information on how warehouse system works such as system performance, ETL job status and system access history.

Data warehousing can be used in organization because it store an aggregation of company data, The main advantage for organisation to maintain efficiency, profitability and Competitive advantages, In organization data is collected through many sources such as online, call centres and inventory management. These data have different degree of relevance and important for companies. Data collected through pass through “conveyor belt” called Data life cycle management.


A data warehouse maintains a copy of information from the source transaction systems, it also maintain data history by integrating data from multiple source systems to be in at central place at the time of need. This is highly beneficial when organization is very big in nature.It improves data quality by encrypting and coding and cleansing the faulty data, maintain data in common format that makes easy to interpretate . By restricting the data it deliver excellent query performance, add value to operational business applications especially customer relationship management. The data ware housing has different design methodologies like bottom-up, top- down & hybrid design. In bottom-up approach data mart are first created to provide reporting and analytical capabilities for specific business process. Data mart contains, primarily, dimension & facts. The integration of data marts in the data warehouse is centred on the conformed dimension-residing in the bus, that possible integration “points” between the data marts, If integration via bus is achieved, it will able to deliver the specific information but also the cumulative one also. In top-down approach the data warehouse is designed using a normalized enterprises “Atomic” data model. Data warehouse is subjected-oriented, all data element relating to the same world event are linked together. It is also Non- volatile in nature, never over written or deleted, data stored is integrated and time variant in nature. In Hybrid design method allow data warehousing to be replaced with “ master data management”, solution where operational, not static information could reside.

Evolution of data warehousing in organisational use takes many steps such as “Offline operational data warehouse “ in which update are maintained on cycle basis such as weekly or monthly. “On-time data warehouse” in which real time data is integrated, “Integrated data base” in which data is integrated from different sources. Operational systems are designed for data integrity, Operational system designers generally follow the Codd rules .Relational databases are clever at managing the relationships between these tables and different data formats. Finally, in order to improve performance, older data


are usually periodically purged from operational systems. Dimensional v/s Non- dimensional approach-In dimensional approach transaction data are partitioned into “Facts” (numeric transaction data) and “Dimension” (reference information). The main disadvantages of this is complicated loading system, due to integrity of facts and dimension, it is very difficult to modify the data warehouse when business changes the way. In normalized approach, divides data into entities, which creates several tables in a relational database. The main disadvantage of this is integration of data from different database into single platform.

The main advantage of data warehousing is for organisation is the simplicity of data for analysis purpose, companies can handled public query easily, implement its operational plan and communicate all plan to workers. It also help organisation to do grow better through efficient management. It help in doing financial analysis and preparing final budget and fulfilling its commitment. Examples –According to a market research company “The Palo Alto Management Group Inc.”, indicate worldwide spending on data warehousing hardware, software, services and inhouse development is increasing by more than 50 percent per year between now and 2002. Top executives of United Parcel Service of America Inc to continue increasing investments in data warehousing hardware and software. . As the $22.5 billion company struggled to


recover from one of the most devastating labour strikes in its history, UPS marketing officials using 400GB data warehouse, to access the company's Information Library, UPS officials across the country were able to quickly target key customers and encourage them to come back to UPS after the strike and work together . Last January, just five months after the strike was settled, UPS was having to shipping as many as 17.2 million packages per day--well above the company's official estimations. For many large companies such as UPS, getting handsome payback from data warehousing has not come quickly or easily. UPS, for example, started its data warehousing efforts in 1994 with large, companywide data quality and standards campaign. Prominent examples of success of data ware house is American Airlines, which at one time is struggling to check the fraudulent tickets which is degrading the revenue of company, after using data ware house system its become easy to track sold ticket. Another glittering example of success of data warehousing is launching of new lines of greeting card names expression by hallmark, which was the success by proper monitoring of revenue data.


ReferencesData warehousing architecture Retrieved from, “” Data design approach Retrieved from, “”

Data warehousing examples Retrieved from, “”

Master your semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master your semester with Scribd & The New York Times

Cancel anytime.